Tuberculosis, one of the oldest human diseases, still kills approximately 3 million people annually, despite the use of a generally safe, but sometimes ineffectual vaccine. While this staggering number is frightening enough, 1/3 of the world s population has been infected by the causative organism, Mycobacterium tuberculosis. Depending on the immunological state and/or age of the latently infected individual, the disease can reoccur, without new infections, at a significant frequency. Antibiotics can control tuberculosis, when therapy is given quickly and efficiently, but the disease is not cured and quiescent M. tuberculosis can be reactivated. Antibiotic resistance is also a growing problem. Currently, little is known about bacterial survival mechanisms in the lung and the nature of the so-called latent state. The general aims of this new grant proposal are: the identification and characterization of M. tuberculosis genes that are specifically induced late in the infection of animal models. Information concerning these differentially regulated bacterial genes and their products will help us to understand interactions between M. tuberculosis and cells in the lung. This knowledge will aid in the development of new anti-tubercular therapies and will provide bacterial markers that will be useful in evaluating the efficacy of other treatments. An IVET romoter trap system has been previously developed in our laboratory to isolate M. tuberculosis genes that are induced after infection of human macrophages and mice. We will extend the IVET screening method to guinea pigs, because the progression of the disease in lungs of this animal model more closely resemble the human disease than that observed in mouse infections. In addition, we have modified our mouse IVET protocols to identify M. tuberculosis genes that are expressed late in the infectious process. The expression of M. tuberculosis genes in infected lung tissue will also be measured. A sensitive and accurate fluorescence based RT-PCR method has been developed to quantitate the in vivo expression of M. tuberculosis genes. Several genes that are induced during infection of human macrophages have already been identified including several iron regulated genes. We will extend this technique to study bacterial gene expression in guinea pig granulomas and mouse lungs. M. tuberculosis genes identified by the experiments outlined above as being induced late in mouse or guinea pig infection will be tested for their roles in virulence. Using techniques developed in our laboratory and others, we will systematically inactivate these genes and will test the ability of mutants to survive in macrophages and animal models.